How to Process a Large Number of Objects

Samples > Rule-Based Programming > How to Process a Large Number of Objects

How to Process a Large Number of Objects

This sample shows how ILOG JRules can be adapted or configured to fit some cases where performance is crucial. The scenario executes a simple cashback computation process involving accounts and a very large number of credit card events per account.

Key Features

Key features include:

Finders: The FinderFinancialEventFactoryImpl.findByAccountId(String accountID) finder is called by the collect condition of some rules. Its purpose is to return a list of correct objects (in the case of this sample, the events attached to an account) and prevent exhaustive and costly iterations required to compute join conditions on large numbers of objects. A finder generally enables you to avoid inserting these objects into the engine's working memory.
Collectors: The CashEventCollector extends IlrDefaultCollector which is used in the collect condition to carry the collection to be returned. This specific collector publishes additional methods (getTotalAmount and getAverageAmount) available in some rule conditions. This allows these computations to be shared at the condition level in the RetePlus and then prevents them from being executed in the actions of several rules.

Importing this Sample

To import the manyobjects sample:

Click File > Import.
In the Import Wizard, select ILOG Rule Studio Samples and Tutorials and click Next.
Select the samples > engine > manyobjects sample.
Click Finish.
If one or more files already exist in your workspace, a warning message is displayed indicating that the corresponding file(s) were not re-imported. Close the message window.

The imported sample project files are displayed in the Rule Explorer.

Running this Sample

Make sure you are in the Rule perspective before running this sample.

To run the sample:

Click Run > Open Run Dialog.
Select the Java Application With Rules > manyobjects-application launch configuration.
Click Run.

Output is displayed in the Console view.

Highlights

The rule project manyobjects-rules contains the ruleset to be executed. An ilrmain function creates and inserts objects in the working memory and executes the ruleset. The rule execution output is displayed in the Rule Studio Console.

An engine.conf file is used to activate the useFinders option to assist with rule execution.

For each account to be processed:

A cashback event is created if there is at least one credit card purchase and the account balance is positive.
Rates are calculated depending on the average of purchases on this account.
A REFUND event is created with an amount depending on the rate and the total amount of the credit card purchases. Note that the sample stops here and does not process the refund.

This sample uses the flexibility of ILOG JRules to solve the problem. It provides ways to perform efficient pattern matching on many objects using ruleflow, finders, and collectors.

The rule project contains:

the Technical Rules
the parameter declarations
the ilrmain function that creates objects for simulation, and then executes for each account object

Rules are applied on a set of instances of several classes. This object model is a simple model made of few interfaces, implementation classes, and factories. It is referenced from the financialevent-xom Java project.

At runtime, the sample processes all the created accounts and logs information on the console. A large number of cash events are processed in a short time, and accounts are granted a REFUND computed with the rule-based rate. By default, 50,000 events are processed on 20 accounts (giving 1,000,000 possible combinations). The Java application provided in this sample is pre-configured to raise the maximum Java Virtual machine memory in order to successfully process this many combinations.